Inspiration
It all started with a common struggle: a guy who could never quite capture photos of his girlfriend that met her expectations. Each photoshoot ended in disappointment and frustration.
What it does
BF Saver uses Gemini AI to analyze and score photos in real-time, providing distinct, actionable advice for both the photographer and the subject. This MVP:
- Identifies the surrounding environment and recommends three curated poses tailored to the scene
- Generates reference images for users to follow
- Provides specific guidance on camera angles and subject positioning
How we built it
Our development process followed these key stages:
- Initial Design: Manual sketching of user flows, followed by UI design in Figma
- Development Environment: Exported Figma code to Antigravity/Cursor as our main IDE
- AI Incorporation: Extensive fine-tuning of Gemini's responses to generate helpful, contextual advice for both posing and shooting techniques
The core challenge was teaching the AI to act as an effective photography director, translating complex aesthetic principles into simple, actionable guidance.
Challenges we ran into
Tool Limitations: No single platform could embody all our requirements. We combined multiple development tools and AI assistants throughout the project.
Background Realism: Initial versions generated posing suggestions with virtual backgrounds that didn't match real scenarios. We refined our prompts to ensure suggestions always reflected the actual environment.
Pose Matching Algorithm: Balancing precision with usability proved critical—the matching needed to be accurate enough to be helpful, but forgiving enough to remain user-friendly rather than frustratingly strict.
UI/UX Optimization: Ensuring seamless navigation and reliable data persistence across screens required significant iteration.
AI Consistency: Maintaining high-quality, contextually accurate evaluations and suggestions required rigorous prompt refinement.
Accomplishments that we're proud of
The MVP received positive feedback from early testers. Users found the scoring system engaging and appreciated the AI's direct feedback. These reviews have been encouraging and give us confidence as we work toward the launch version.
What we learned
Beyond technical challenges, we gained practical insights into photography fundamentals. We learned that aesthetics are highly subjective with many valid interpretations. Converting this subjectivity into standardized algorithms that provide clear, helpful guidance was significantly more challenging than anticipated.
What's next for BF Saver
Immediate priorities:
- Refining the frontend based on initial user data and feedback
- Optimizing for different mobile operating systems to ensure cross-platform compatibility
- Improving overall user experience flow
Long-term goals:
- Developing a more intuitive design for photography guidance and instructions
- Collecting comprehensive user insights for final polish
- Launching the app
Log in or sign up for Devpost to join the conversation.